A pulse doppler radar system for helicopter recognition has a receiver (5) for receiving reflected pulse radar signals. A distinguishing circuit (22) is provided for distinguishing a group of received frequencies within a predetermined spectral range and for ignoring at least the largest amplitude frequency within that group. A threshold detector (24) is arranged to determine when the amplitude of the remaining frequencies in the group exceed a predetermined value to provide a recognition signal.
|
1. A pulse doppler radar system for helicopter recognition, comprising a receiver for receiving reflected pulse radar signals, means for distinguishing a group of received frequencies within a predetermined spectral range and for ignoring at least the largest amplitude frequency within that group, threshold detection means for determining when the amplitude of the remaining frequencies in the group exceed a predetermined value and for providing a recognition signal and an accumulator for combining the signals of said remaining signals to provide an aggregate amplitude for evaluation by the threshold detection means wherein the threshold detection means has a series of weighted threshold values indicative of different levels of certainty of helicopter recognition.
2. A system as claimed in
3. A system as claimed in
4. A system as claimed in
5. A system as claimed in
6. A system as claimed in
8. A system as claimed in
|
The invention relates to pulse doppler radar systems and more particularly to the recognition of helicopters by such systems.
Conventional pulse doppler radar systems experience difficulty in detecting slow moving objects because they are arranged to filter out reflections from stationary and slow moving objects, generally known as radar clutter. It would be advantageous however to be able to detect helicopters which with current systems are ignored. The present invention is concerned with such detection.
It has been observed that pulses returned by helicopters in doppler radar systems display spectral spreading through a FFT (fast fourier transform) system. The spectral spreading is due to vibration and to rotation of the hub and is not experienced in the reflections from other stationary or slow moving bodies, although all moving bodies will provide a narrow spectral component corresponding to their radial velocity. In an eight point FFT system this corresponds to a response in one filter, indicative of radial movement, sometimes with sidelobes in one or both adjacent filters. Most moving objects have a response in the remaining filters based only on system noise. Helicopters, however, have response in these filters due to the amplitude of the radial velocity component and the spectral spreading.
It was initially thought that if a threshold was set against all filters, e.g. eight in a particular range cell, then a straight "n" out of eight criterion could be set for helicopter identification. As long as the threshold is set above noise, doppler reflections due to normal slow moving objects, would have amplitudes which would cross the threshold only a maximum of three out of eight. It was thought that if "n" was set to four, this should result in helicopter identification whilst minimising false alarms from other bodies. There are a number of basic problems with this method. Firstly, it is possible for other objects to induce some minimal spectral spreading, due for example to vibration. This necessitates a value of "n" greater than four to reduce false alarms. Secondly, with this larger value of "n" the threshold may have to be lowered as the distribution of the helicopter spreading in the remaining filters may not include two above the threshold (this assumes an "n" value of five with the three radial velocity filters obviously greater than the threshold). The result of this is that the thresholds are very difficult to optimise for sensible helicopter identification whilst minimising false alarms.
The previously described method due to its amplitude thresholding throws away a lot of useful information. If the response due to the target radial velocity is removed, the remaining filters will contain information due to spectral spreading only. It is upon this consideration that the present invention is based.
According to the invention there is provided a pulse doppler radar system for helicopter recognition, comprising a receiver for receiving reflected pulse radar signals, means for distinguishing a group of received frequencies within a predetermined spectral range and for ignoring at least the largest amplitude frequency within that group and threshold detection means for determining when the amplitude of the remaining frequencies in the group exceed a predetermined value and for providing a recognition signal.
The invention effectively ignores at least the largest amplitude frequency, as this relates to the radial velocity of the body. This allows the smaller spectral responses to be assessed and compared with a threshold value to enhance the sensitivity and accuracy of helicopter detection.
As previously mentioned some objects other than helicopters, can produce spectrum responses which extend into the two filters immediately adjacent the filter which passes the largest amplitude frequency. Therefore in a refinement of the invention, the means for distinguishing a group of received frequencies within a predetermined spectral range is arranged to ignore the largest amplitude frequency and each immediately adjacent frequency in the group.
In order that the invention and its various other preferred features may be understood more easily, an embodiment thereof will now be described with reference to the drawing which is a simplified schematic block diagram of a pulse doppler radar system incorporating a helicopter recognition system constructed in accordance with the invention.
In the drawing, an input signal from a receiver 5 is coupled via a fast fourier transform circuit (FFT) 10, in this case an eight point FFT is employed. The output of the FFT comprises eight multiplexed filter channels and is fed to a set of cell-averaging circuits 11 which implement the CFAR function. The output of the cell averaging circuits 11 is routed to a target detection system comprising a maximum amplitude signal detector 12, a threshold comparator 13 coupled with the output y1 of the signal detector, which compares the signal y1 with a number of differently weighted threshold values, for example 3 T15, T35, T55, a summing circuit 14, a progressive storage circuit 15 and second threshold comparator 20. The output of the threshold comparator 13 which is a weighted value of 0, 1, 3 or 5 is fed to one input of the summing circuit 14 and also to the first stage of the progressive storage circuit 15 which has a series of memory banks 16 to 19 through which successive sample weighted values from the threshold comparator are sequentially advanced. The value stored in each memory bank is also coupled with an input of the summing circuit which provides an output which is fed to one input of the second comparator 20. The second comparator compares the aggregated signal from the summing circuit 14 with a threshold value T25 and when it exceeds that threshold it provides an output indicative of the presence of a target.
The helicopter recognition system is shown at 21 and comprises a signal amplitude detector and discriminator 22 an accumulator 23, a threshold comparator 24, a progressive storage circuit 25 which has a series of memory banks 28 to 31, a summing circuit 26 and a second threshold comparator 27.
The signal amplitude detector and discriminator 22 determines the largest amplitude received frequency and sets the weighted amplitude value of this and the immediately adjacent frequencies to zero thereby effectively ignoring these responses and providing outputs relating only to the remaining frequencies, in this embodiment five frequencies. These five amplitudes in themselves are each indicative of the presence of a helicopter. However, to provide increased certainty of detection, and avoidance of false detection, the amplitudes are fed to the accumulator 23 and the accumulated value is fed to the threshold comparator 24 where it is compared with a series of weighted values. In the described embodiment three weighted values are employed.
The output of the threshold comparator 24, which is a weighted value of 0, 1, 3 or 5, is fed to one input of the summing circuit 26 and also to the first stage 28 of the progressive storage circuit 25, successive sample weighted values from the threshold comparator are sequentially advanced through the memory bank stages 28 to 31. The value stored in each memory bank is also coupled with an input of the summing circuit 26, which provides an output which is fed to one input of the second comparator 27. The second comparator compares the aggregated signal from the summing circuit 26 with a threshold value T2H and when it exceeds that threshold value it provides an output indicative of the presence of a helicopter.
By employing the remaining five filter signals and accumulating their amplitudes, the large spectral spreads, due to helicopters, are easily distinguished from the minimal amount seen on some doppler targets. To achieve this, the accumulated value is compared to a set of thresholds weighted 1, 3 and 5. If the accumulated value is less than all thresholds, it is assigned a weight of zero, a miss. If it is greater than 1.1 assign a weight of 1, if greater than 1.3 a weight of 3, if greater than 1.5 a weight of 5. For a particular range cell, the current weight plus the weights from four previous processing intervals (corresponding to an antenna beamwidth) are summed. This is a standard multi-level moving window integrator function. The sum (generated over a beamwidth) is now compared to a second threshold to determine if a helicopter is present of not. Trial results have shown that the weighted first threshold and second threshold can be chosen to maintain very good target identification of a helicopter, whilst minimising false identification of "normal" doppler targets, even when they exhibit a certain amount of spectral spreading.
Although the embodiment described employs an eight point fast fourier transform system, it will be appreciated that the invention is not restricted to eight point systems. Other systems e.g. sixteen point fast fourier transform systems can be employed.
Although the described embodiment effectively ignores the maximum amplitude signal and immediately adjacent frequencies to cover side lobes, it is envisaged that in certain circumstances an operational system may be achieved ignoring only the maximum amplitude signal.
Instead of employing the CFAR function, the block 11 could otherwise be implemented as parallel channels up to the selection function 12.
Lewis, Christopher, Barton, Robert J.
Patent | Priority | Assignee | Title |
10514454, | Oct 24 2014 | The United States of America, as represented by the Administrator of the Federal Aviation Administration | Techniques for mitigating the effects of complex structures on radar systems |
5280289, | Apr 23 1992 | HE HOLDINGS, INC , A DELAWARE CORP ; Raytheon Company | Automatic signal thresholding system |
5311188, | Dec 19 1991 | THALES NEDERLAND B V | Radar apparatus provided with a coherent clutter map |
5347282, | Oct 16 1991 | THALES NEDERLAND B V | Apparatus for the observation and indentification of helicopters |
5689268, | Aug 02 1996 | BOEING NORTH AMERICAN, INC | Radar detection and classification of helicopters |
7522089, | Jun 12 2006 | Raytheon Company | Airborne look-down doppler radar tracking of hovering helicopters using rotor features |
9041592, | Oct 14 2011 | HANWHA VISION CO , LTD | Radar sensor and method of detecting object using the same |
9870620, | Jun 18 2015 | The Johns Hopkins University | Non-linear associator and motion discriminator |
Patent | Priority | Assignee | Title |
4275396, | Oct 12 1979 | Northrop Grumman Corporation | Helicopter rotating blade detection system |
4348674, | Jul 18 1979 | Siemens Aktiengesellschaft | Apparatus and method for classifying moving targets |
4389647, | Dec 22 1980 | The United States of America as represented by the Secretary of the Army | Doppler discrimination of aircraft targets |
4641138, | Nov 06 1975 | Lockheed Electronics Co., Inc. | Radar apparatus for detecting an agitated reflective target |
DE3002148, |
Executed on | Assignor | Assignee | Conveyance | Frame | Reel | Doc |
Nov 08 1990 | Siemens Plessey Electronic Systems Limited | (assignment on the face of the patent) | / | |||
Dec 11 1990 | LEWIS, CHRISTOPHER | Siemens Plessey Electronic Systems Limited | ASSIGNMENT OF ASSIGNORS INTEREST | 005571 | /0549 | |
Dec 11 1990 | BARTON, ROBERT J | Siemens Plessey Electronic Systems Limited | ASSIGNMENT OF ASSIGNORS INTEREST | 005571 | /0549 | |
Mar 26 1991 | Plessey Overseas Limited | Siemens Plessey Electronic Systems Limited | ASSIGNMENT OF ASSIGNORS INTEREST | 005668 | /0201 | |
Sep 12 2003 | Siemens Plessey Electronic Systems Limited | BAE SYSTEMS DEFENCE SYSTEMS LIMITED | CHANGE OF NAME SEE DOCUMENT FOR DETAILS | 014683 | /0875 |
Date | Maintenance Fee Events |
May 15 1995 | M183: Payment of Maintenance Fee, 4th Year, Large Entity. |
May 12 1999 | ASPN: Payor Number Assigned. |
May 17 1999 | M184: Payment of Maintenance Fee, 8th Year, Large Entity. |
May 23 2003 | M1553: Payment of Maintenance Fee, 12th Year, Large Entity. |
Jun 18 2003 | REM: Maintenance Fee Reminder Mailed. |
Date | Maintenance Schedule |
Dec 03 1994 | 4 years fee payment window open |
Jun 03 1995 | 6 months grace period start (w surcharge) |
Dec 03 1995 | patent expiry (for year 4) |
Dec 03 1997 | 2 years to revive unintentionally abandoned end. (for year 4) |
Dec 03 1998 | 8 years fee payment window open |
Jun 03 1999 | 6 months grace period start (w surcharge) |
Dec 03 1999 | patent expiry (for year 8) |
Dec 03 2001 | 2 years to revive unintentionally abandoned end. (for year 8) |
Dec 03 2002 | 12 years fee payment window open |
Jun 03 2003 | 6 months grace period start (w surcharge) |
Dec 03 2003 | patent expiry (for year 12) |
Dec 03 2005 | 2 years to revive unintentionally abandoned end. (for year 12) |